Unless you want to model a thin film in vacuum, you should NEVER let all atoms
relax in a surface calculation. In reality, the lowest layer(s) are attached to
the crystal that's much thicker than your model slab, so you have to freeze them
(or at least one) at the bulk geometry.
The only possible argument for full relaxation is that you want to prevent an
artificial dipole moment from the fact that the upper surface relaxes and the
lower one doesn't, but even then I would freeze the middle layer.
Herbert
On 05/12/11 10:02, Ary Junior wrote:
Hi,
Increase the number of layers and let all atoms to be relaxed... So you will
check "how deep" the relaxation of the atoms from surface will influence the
atoms of the bottom layers... With that test you will have the correct slab
tickness (which will be the half of the converged slab found in the test) and
number of layers to be relaxed in your simulations...
Ary Junior
Universidade Federal de Juiz de Fora - MG - Brasil
On Sun, Dec 4, 2011 at 4:46 AM, mehzad fatehan <[email protected]
<mailto:[email protected]>> wrote:
hi thanks
energytotal - number of layers ?
On Sat, Dec 3, 2011 at 7:31 PM, Ary Junior <[email protected]
<mailto:[email protected]>> wrote:
Hi,
Do a convergence test and you will find the suitable values for the slab
tickness and number of layers to be relaxed.
Ary Junior
Universidade Federal de Juiz de Fora - MG - Brasil
On Sat, Dec 3, 2011 at 1:58 PM, mehzad fatehan <[email protected]
<mailto:[email protected]>> wrote:
hi every body
i am new user
i want to know Suitablenumber of layers in surface, please help me
very thanks
--
http://lattes.cnpq.br/8221674673413336
--
http://lattes.cnpq.br/8221674673413336
--
Herbert Fruchtl
Senior Scientific Computing Officer
School of Chemistry, School of Mathematics and Statistics
University of St Andrews
--
The University of St Andrews is a charity registered in Scotland:
No SC013532
